The milecastles are fortlets attached to the curtain wall which, it is assumed, were used for outposting troops from the forts and as bases for patrolling the curtain wall and ditch. Of something like 81, most are not visible (some were even removed by the Romans. Those that can be seen range from the fully excavated to humps and bumps in the ground, but they are all more or less the same size.
Milecastle 50 TW (High House) [HB 309–12]
To say that there is not much to see of Milecastle 50 TW (High House) is probably something of an understatement but no less true for all that. It has the distinction of being the only Turf Wall milecastle without a Stone Wall successor on top of it. Excavated in 1934, it was found to have an undug causeway across the ditch, a rampart of turf and gateways of timber (which, it is suggested, had towers, since there were thought to be too many timbers just for revetting the gateway). The Vallum ditch swerved south to avoid the milecastle, but more interestingly the excavations found a timber inscription, restored as recording its construction under A. Platorius Nepos. It is customary to think of Roman inscriptions as being carved in stone, but (as on the milestone in Wall Mile 43) they could be painted on stone or carved (or painted) on wood. These are amongst the Rumsfeldian ‘known unknowns’ of Roman archaeology.
Harrow’s Scar Milecastle (MC49)
Much of the interior of Milecastle 49 (Harrow’s Scar) has been removed by the modern track which passes through it, but this is the first consolidated milecastle encountered when walking from west to east. It therefore provides our first real opportunity to get the measure of one of these fortlets, although it is not as informative as its neighbour, Milecastle 48. The main structure inside the milecastle is part of a medieval farmstead, recalling just how many milecastle sites came to serve as a farm. On the east side there is one wall of an original internal Roman building. The rounded south-west corner is well-preserved. Astute observers will note how the defensive walls of the milecastle butt against the curtain wall. This tells us that the curtain wall was built first, then the milecastle. The modern farm track does not use the Roman gateway, which is just to the west of it.
Poltross Burn Milecastle (MC48)
Milecastle 48 is one of the best preserved along the line of the Wall. First excavated in 1886, and then subsequently in 1909 and 1911 (and again in 1965 and 1966), it is perched rather precariously behind and sloping down towards the curtain wall. Its own perimeter wall joins the curtain perpendicularly (and is bonded to it), but its southern corners are rounded as Roman fortifications of the 2nd century AD usually were. On the eastern side (and on the western, but no longer visible) is a short wing wall similar to those on Turrets 48a and 48b, confirming that the milecastle had been built to the broad gauge before the (now narrow) curtain wall reached it. Here it differs radically from its later neighbour, Milecastle 49, which, as we have seen, was built after the curtain wall.
There are two internal structures, aligned on either side of the roadway that passes between the gates, the northernmost of which, rather excitingly, led out into Barbaricum (these days represented by the side of the embankment of the Newcastle to Carlisle railway). The north gate is well preserved and is of a type (III, if you must know) that scholars have suggested mean it was constructed by legio VI Victrix. The gate was subsequently narrowed to only allow pedestrian access, as happened at other milecastles.
Within the milecastle, a north–south road way ran between the north and south gates and on either side of it were structures thought to be barrack buildings. The slope is rather extreme and the floors were probably levelled up on joists inside. It is thought that the milecastle garrisons were outposted from neighbouring forts, in order to provide the manpower to patrol the Wall and man the turrets. ‘Thought’ because, in the absence of direct evidence from Hadrian’s Wall itself, we have to use comparisons with other provinces, where we know outposting was practised widely.
In the north-east corner are the remains of the oven used for cooking (the Roman army preferred to keep their celebrity chefs at arm’s length), so it is in the equivalent position to ovens in larger forts and fortresses. Roman ovens worked like a traditional pizza oven: the fabric is heated by inserting hot embers which are then raked out and the food to be cooked (mostly bread) placed inside. Carbonised Roman bread (as well as wall paintings of the uncarbonised original) is known from Pompeii, but as yet no pizza.
Milecastle 48 also provides another important fact that sheds some (but not quite enough) light upon our understanding of the Wall. This is the survival of a staircase in its north-east corner. Only the lowest three (and part of a fourth) steps actually survive, but by projecting their line upwards, it is possible to deduce that the wall-walk height here must have been in the region of 12 Rft (3.55m). The astute observer will note that the stairs were constructed after the north and east walls of the milecastle, since it butts against and is not bonded with them (the coursing of the facing stones is different).
This then leads on to the question of whether the curtain wall had a walkway along the whole length of its top, or just in select places, such as round the perimeter of the milecastle defences. Scholars can be found who favour either; the evidence is suitably and intriguingly ambivalent; but, in the end, it all comes down to a matter of personal preference. Those brought up on a diet of Alan Sorrell illustrations, with a crenellated Wall striding across the crags, wind-blown Roman soldiers atop it warily eyeing distant squalls, find it hard to escape the image, even though we know other contemporary frontiers (notably those in Germany) had no such feature and modern military walls (like the Berlin Wall or the coyly named Security Fence around the West Bank in Palestine) get by quite happily without being patrolled from the top. No certainties there, then, but there are many who would argue that the width of Hadrian’s Wall only makes sense if there was a walkway.
Walltown Milecastle (MC45)
Milecastle 45 (Walltown) has never been properly excavated, but the robber trenches dug to remove its walls (and their associated spoil heaps) are very clear and serve to delineate the structure. Comparison with Milecastle 48 next to the Poltross Burn shows the two extremes of preservation and presentation.
Allolee Milecastle (MC44)
The long-axis Milecastle 44 (Allolee) is visible as an earthwork. It has evidently been excavated at some point but when, by whom, and what was found remain a mystery. This is why excavation is sometimes euphemistically referred to as ‘preservation by record’; without a record, there is no ‘preservation’.
Cawfields Milecastle (MC42)
Milecastle 42 was first excavated by John Clayton in 1848, re-excavated in 1935, and finally consolidated in the 1960s. The outline of this somewhat drastically sloping short-axis milecastle is clear but no internal structures were identified (or survived). The gateways are rather noteworthy, the south for the height of its masonry, the north for its marking-out lines (used to align its now-missing masonry) and for the remaining sockets for its gates. A fragment of tombstone and a Hadrianic building inscription come from here.
Melkridge Milecastle (MC41)
Milecastle 41 (Melkridge) is a short-axis milecastle that survives as humps and bumps. The site was subsequently occupied by Shield on the Wall farm, which was later relocated 500m to the south-west, just south of the Vallum.
Winshields Milecastle (MC40)
Unusually, Milecastle 40 has been placed at an angle in the curtain wall. It is easy to locate by the gap in the field wall on top of the curtain wall. It was excavated in 1908 and found to be a long-axis type with both gateways reduced to pedestrian-only access at a late date. What we see today is the bumpy aftermath of those excavations.
Castle Nick Milecastle (MC39)
The surrounding walls of the long-axis Milecastle 39 (Castle Nick) are Claytonized and it has been the object of the attentions of excavators in 1854, between 1908 and 1911, and most recently in the 1980s. The structures visible inside it are for the most part post-medieval (it is said to have been used as a milking parlour) and demonstrate once again the re-use of milecastles for agricultural purposes in later years. It was not located in its measured position but further east, perhaps deliberately to place it in the gap.
Hotbank Milecastle (MC38)
Milecastle 38 (Hotbank) is a short-axis site with type I gateways, the northern of which was narrowed to pedestrian access only. The site is significant in many ways, both from a historical and conservational standpoint. It is the origin of one (RIB 1638) and probably two (RIB 1637) of the building inscriptions that confirm that the builder was A. Platorius Nepos on behalf of Hadrian, thereby showing that Hadrian’s Wall really was… Hadrianic! There is the chance to inspect one of these in the Great North Museum in Newcastle.
Although its robbed remains were excavated in 1935, it is now presented only as an earthwork and one that provided a nasty (if Pythonesque) moment for those charged with the upkeep of Hadrian’s Wall. In 2003, a group of 850 Dutch bankers on a team-building ‘jolly’ visited it in damp conditions and unintentionally caused considerable (and of course irreparable) damage to the earthwork. However, lessons on the management of the monument were learned from this and subsequent visits by massed murophiliac (and presumably high-fiving) bankers have been handled much more successfully. All of which means there is a very good reason the path goes round Milecastle 38, not over it.
Housesteads Milecastle (MC37)
Milecastle 37 is perhaps the most visited, by dint of the fact it is closest to Housesteads (which enjoys the highest visitor numbers for Hadrian’s Wall), and is within staggering distance for the more adventurous car-bound visitor. It is presented in the same Claytonized form as the curtain wall on either side, facing stones reconstructed up to a regular height and topped with turf. It has been excavated four times between the middle of the 19th and end of the 20th centuries and, quite apart from offering an excellent sheltered location for a walker’s lunch, provides our first insight into the nature of the milecastle.
Beginning with the north gate, we can see that the reduction in width to pedestrian access is still in place. Comments are occasionally made that it is daft to provide gateways for some of the milecastle along the crags, but access would have been needed along the front of the curtain wall and ditch for the purposes of maintenance and many afforded some sort of rudimentary route to the north, the pedestrian blocking being a recognition of the fact that this was probably usually not by wheeled vehicle. In fact, the most recent excavation showed how partial collapse of the north gate led to its being blocked soon after construction and only opened up for pedestrian access at a later date. The lowest two voussoirs of the southern arch of the north gate are still in place on either side, but the others have been replaced in recent times for effect (a drawing of 1879 by James Irwin Coates shows those two springers, as they are known, in situ).
There is one internal building, east of the central north–south roadway, whilst the only excavated sign of a western structure here was a couple of hearths.
King’s Hill Milecastle (MC36)
Best appreciated for its commanding position, rather than its remains, Milecastle 36 (King’s Hill) was a long-axis example perched on a hill, overlooking Busy Gap to the north-east. The identity of the king in question will become clearer once we get up onto Sewingshields Crags in a short while.
Sewingshields Milecastle (MC35)
Milecastle 35 (Sewingshields) was excavated in 1978–82 and the first thing the visitor notes is that this long-axis milecastle has no north gate. This is one of those few instances where it would be truly superfluous. The interior of the fortification is occupied by several phases of Roman building on either side of the central roadway, culminating in its re-use as a medieval farmstead. The later Roman phases included evidence of metalworking on the site.
Tower Tye Milecastle (MC29)
Milecastle 29 (Tower Tye), like Milecastle 38, exists now solely as an earthwork, but is nevertheless an extremely interesting example. Excavated by John Clayton, the robber trenches for its walls are still sharply defined. However, there is an additional detail that the keen-of-eye may be able to make out and that is the fact that the milecastle is one of the few known to have had a ditch around it. It shows up now as a shallow depression around the west, south, and east sides.
Westgate Road Milecastle (MC4)
Scholarly calculations had long been baffled by the location of this (as it turns out) long-axis milecastle, not least because of the uncertainties we are about to encounter over the course of Wall Mile 3. In 1985, its discovery in the backyard of the Newcastle Arts Centre during the digging of a drain led to Milecastles 5 and 6 being shuffled along a bit from their old hypothetical locations to new hypothetical locations, and all was better. Its position is marked in the Black Swan Yard behind the Arts Centre.
To hear an interview with Dr Matt Symonds, one of the foremost authorities on Hadrian’s Wall milecastles, listen to Podcastellum 8.